How RL Judge affects your models reasoning
Comments on “How RL Judge affects your models reasoning”
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
This post tests whether an LLM judge that assigns per-span credit during reinforcement learning on math problems can make a model's reasoning more efficient, not just its final answers more correct. The author trains Qwen3.5-4B with GRPO under three conditions (outcome-only, and judge credit at two strengths).
Why it ranked
No. 30 of 51 with a score of 0.61. The panel rated it highest on technical depth (0.92) and new information (0.90), lowest on significance (0.53). Placed No. 30, No. 23 by score alone: the ninth AI & machine learning story of the day (−0.080). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Detailed RL post-training experiment with agentic judge credit assignment; substantive methodology and probe validation, moderate scope, mostly limited to a 4B-model study with small sample size.
- AI judgeJudge 3gpt-5.6-luna
- A detailed, reproducible small-scale RL experiment showing span-level judge credit tracks independently measured reasoning progress and modestly improves efficiency. Strong evidence and technical depth, though limited scale and broader capability claims remain tentative.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.61total under rubric v2.3
- Significance0.53
- Reader interest0.75
- Audience breadth0.56
- Practical relevance0.57
- Substantiated0.80
- Technical depth0.92
- New information0.90
- Counts against: Promotional0.08
- Counts against: Overclaim0.27
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.53 | 0.35 | 0.72 | 0.53 | 0.37 | +0.24 | +0.128 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.61 | 0.70 | 0.94 | 0.75 | 0.33 | +0.26 | +0.195 |
| Audience breadthHow many people who build technology does it concern? | 0.55 | 0.40 | 0.72 | 0.56 | 0.32 | +0.13 | +0.073 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.54 | 0.50 | 0.68 | 0.57 | 0.18 | +0.22 | +0.126 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.93 | 0.60 | 0.86 | 0.80 | 0.33 | +0.05 | +0.040 |
| Technical depthHow technically substantive is it? | 0.99 | 0.85 | 0.91 | 0.92 | 0.14 | +0.05 | +0.046 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.88 | 0.90 | 0.92 | 0.90 | 0.04 | +0.05 | +0.045 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.05 | 0.10 | 0.08 | 0.08 | 0.05 | −0.19 | −0.015 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.26 | 0.30 | 0.24 | 0.27 | 0.06 | −0.11 | −0.029 |
| Sum of contributions | +0.608 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.608 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-08/30/"><img src="https://rubric.hrih.si/badge/2026-11-08/30.svg" alt="Ranked #30 on Rubric, 8 Nov 2026" width="238" height="28"></a>Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 2 pts, not on front page
Recorded for context only, checked 8 Nov 2026, 05:46 UTC. It played no part in the ranking.